Don’t Worry, AI Can't Make You Stupid

It’s U.S. Thanksgiving, and many Americans are anxiously running a gauntlet through airports and other transportation corridors. May you all arrive safely at your destinations and enjoy your holiday festivities with family and friends.

Considering the time of year, I could use the occasion to write something uplifting, but that would be far too easy. Instead, I’ll buck convention and go my own difficult way.

Today my focus is on the big-picture question of whether AI – as we come to depend on it to communicate and therefore think on our behalf – will make us stupid. Many minds, great and small, have debated this question. The debate has raged for a while, too, reaching back more than a decade.

Some have argued that we’re at risk of losing our faculties of cognition and intellectual discernment to intelligent automation, while others have countered that the threat is wildly exaggerated. This latter group sometimes cites the precedent of the calculator, which, when it was first introduced, provoked fears that using machines for arithmetic and mathematic calculation would dull our mental acuity.

Proponents of this analogy contend that human intellectual discernment was not impaired by the advent of the calculator; they cite the innumerable scientific and technological breakthroughs that have occurred since then, and that continue to occur with increasing frequency. They argue that humanity continues to wrestle with weighty intellectual concerns. I readily concede that those points are valid.

Nonetheless, there’s a difference between using a calculator and using AI to obviate or reduce the need for human-to-human communication and creative thought. AI is an entirely different proposition, especially artificial general intelligence (AGI), which will follow and evolve from the mimicry of generative AI. If AI gains the capacity to genuinely think and reason, and can successfully deliver cogent communication and correspondence at a level that equals or surpasses that of human attainment, isn’t it possible that many people will push the proverbial “easy button” and let machines perform an increasing array of cognitive tasks on their behalf?

If that happens, what are the personal and social implications? The truth is, regardless of which side of the debate one might be inclined to endorse, we just don’t know the answer to that question. The future truly is unwritten. More to the point, we don’t know who will write it.

Reason to be Cheerful

So far, this post hasn’t offered much hope or joy, but now I will give you a reason to be cheerful. Even though AI, by potentially assuming responsibility for your writing and other communication – and thereby atrophying an essential part of what makes you human – could conceivably dim your cognitive lights, you will retain the power to decide how much, or how little, AI contributes to your intellectual decline.

You needn’t let AI make you mentally indolent. If you decide to use AI, to cite one contemporary example, to draft your correspondence, you can still ensure that you actively and thoroughly read and edit what AI has produced on your behalf. Before you hit send, you can critically assess whether you approve of AI’s prose and modify anything that doesn’t feel or sound right. You can even choose not to use AI at all. It’s entirely within your power, at your discretion. You control what you think and how much, or how little, you want AI to relieve your intellectual burden.

I endorse the view that we should never surrender our duty and privilege of thought to any external party, whether human or machine. Each of us benefits immeasurably from exercising our highly personal powers of thought and expression. What’s more, each of us has a distinct sensibility that is revealed in our thought and expression.

Thought is essential to human existence. What happens if we outsource thought, in whole or in part, to automated intelligence that has been trained on black-box models that are partly or completely opaque? The answers are perhaps unknowable, but the imagined consequences are unsettling.

We already see how social-media algorithms have conditioned highly emotive, often illogical and unreasonable, public discourse. A glance at what’s promoted in most social-media feeds reveals a torrent of bile and effluent. The digital “public square” more often resembles the shambling grounds of a digital madhouse.

Now I know social media is relatively primitive stuff, that future AI will be far more advanced technologically. That’s a blessing and a curse. As now seems likely, the largest purveyors of AI, for both consumers and enterprises, will be extremely large technology concerns, such as Microsoft, Google, Amazon, and Meta. Some startups will develop into major players, but they will necessarily grow into large corporate concerns or be acquired by the industry giants.

If, in the not-too-distant future, you contract the services of an “AI life coach” or an “AI business adviser” from a technology behemoth, you will need to trust that they will put your interests above their own – and perhaps the interests of their advertisers – before your follow their guidance. Does the AI work for you, or for those who trained it and own the capital infrastructure that brought it to commercial life?

A large measure of transparency about the training of AI foundation models, including about the exact data sources and processes that underpin these offerings, would help you make an informed decision regarding how, or whether, to utilize the service. Again, though, you must exercise your capacity for critical thought to make the right decision in your own interests.

As my career attests, I’m not a Luddite. I accept that AI will become a ubiquitous and hugely influential presence in our lives. The proverbial genie is out of the bottle, and there’s no putting it back. The challenge now is to employ the genie prudently and to articulate our wishes wisely, remembering the admonition that one should be exceptionally careful when making wishes.

Big Stakes, Unprecedented Influence

We have just witnessed an incredibly convoluted and melodramatic soap opera at OpenAI. Others have chronicled every serpentine twist and turn in the saga, and I’m not sure we (I’m using the “royal we” to represent all those outside OpenAI) know exactly what happened or why it happened. The narrative has moved past a climax, but it hasn’t reached a denouement, much less an epilogue. It’s hard to say precisely how matters will be resolved.

What I will say with confidence is that you only see such elaborate political intrigue, internecine machinations, and bold gamesmanship in organizations where the stakes are uncommonly high. When inordinate power and vast sums of money are on the line, principals leave their better angels behind and let fear and greed do the driving. The OpenAI implosion makes for grimly fascinating spectacle – like a car crash involving a Bugatti La Voiture Noire – but it hasn’t made for edifying programming. All the principals are exceptionally intelligent people, but their actions in this debacle belie their impressive erudition and qualifications.

We should remember, however, that there’s a lot on the line. AI is not only big business – seemingly destined to become the biggest of all technology businesses – but also one that is poised to confer extraordinary power. After all, if the public and business customers unquestioningly accept AI advisory services as the digital era’s evolved answer to the Oracle of Delphi, they will be inclined to trust it implicitly and unconditionally. If you own and control the intellectual property that animates and informs the most popular AI services, you have power of a rare kind. I would even go so far as to contend that power of that sort is unprecedented.

In closing, let’s come back to the question that I have sought to address, and which has been debated for a decade or more: Will AI make us stupid? No, only we can make ourselves stupid. The answer is in your hands, and in your own powers of cognition. AI can’t make you stupid. If you retain and exercise your own freedom of creative and critical thought, no external force can make you stupid.